agentready-store-audit

AgentReady store audit

Score, out of 100, how well an AI shopping agent can find, understand and buy on an e-commerce store — plus the top fixes. Free, keyless, deterministic (same store → same score). Service: https://www.agentready.market

When to use

• The user asks "is my store ready for AI shopping agents?" or wants an agent-readiness / AOO audit of a shop.
• You are about to attempt a purchase on a store on the user's behalf — check it first; a "capped" store blocks agents and the attempt will fail.
• The user wants to compare stores or find one agents are known to work on.

Not for: news/SaaS/content sites (this audits e-commerce transactions), or "does ChatGPT cite my brand" questions (that's discovery/AEO — different job).

How

Preferred — MCP (Streamable HTTP, no auth): register https://www.agentready.market/mcp and call:

• check_merchant_readiness(host) — score, grade, capped status, protocol adoption (UCP/A2A/ARD) for any already-scanned host. Never triggers a new scan of a third-party site.
• get_readiness_report(host) — the full per-check public report.
• list_agent_ready_stores(vertical?, limit?) — the Agent-Ready Index (grade B+ stores, ranked): where agents succeed.

REST fallback (also keyless):

# Fresh scan of the USER'S OWN store (~30s, synchronous)
curl "https://www.agentready.market/api/v1/scan?url=STORE_URL&vertical=general"
# Read a stored report
curl "https://www.agentready.market/api/v1/scan/SCAN_ID.json"

Interpreting the result

• score /100 and grade A–E: A/B = agents can buy (Index-listed); C/D = partial; E = not usable by agents.
• capped: true = an access hard-block (robots/bot-wall) caps the whole score — agents are blocked at the door; this is always fix #1.
• top_fixes = highest-impact remediations, each with impact points and effort; relay them in order.
• neighbors = 2 stores above / self / 2 below on the Index — give the user a concrete score to beat.
• report_url = shareable human report; include it in your answer.